Zahn, R., Comas, M.C., and Klaus, A. (Eds.), 1999Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 161

8. MINERALOGY AND MICROFABRIC OF SEDIMENT FROM THE WESTERN MEDITERRANEAN SEA1

Jane S. Tribble2 and Roy H. Wilkens2

stes ofents.

nturialoran theritalith

es ofitesiit

ABSTRACT

Analysis of bulk mineralogy and microfabric of sediments collected during Ocean Drilling Program Leg 161 indicates anotably uniform Neogene/Quaternary history of hemipelagic sedimentation in the Western Mediterranean Sea. Bulk X-ray dif-fraction data were collected for 857 samples from six drill sites in an east–west transect from the Tyrrhenian Sea to the wernAlboran Sea. Semiquantitative relative abundances of clay minerals, calcite, quartz, and feldspar are reported. Age profilethe major phases are used to describe downhole and east–west variations in mineralogy of the dominant hemipelagic sedimAlthough overall mineralogy is remarkably similar from site to site, some spatial variability was noted. Calcite conteincreases from an average of ~20% in the westernmost Alboran Sea to ~40% in the Tyrrhenian Basin. Calculated calcite bfluxes correlate with sedimentation rates and define an opposite trend, with greatest accumulation rates in the western AlbSea. Quartz concentrations are slightly higher in hemipelagic sediments from the westernmost Alboran Sea sites than ineastern Alboran, Balearic, and Tyrrhenian Basins. These compositional trends are thought to reflect the influence of detinflux to the basins. Age profiles of calcite flux reveal peaks, many of which correlate between sites. Climate change, wassociated changes in delivery of detrital material to the basins, and possibly in productivity, are likely responsible for the peaksin calcite accumulation.

Backscattered electron images of 36 samples of hemipelagic sediment were collected and analyzed for areal percentaggrains, macropores, and matrix. Microfabric samples from a depth profile at Site 974, and across three time slices at all s,show only minor variations in gross fabric related to age, depth of burial, or location. All samples are matrix dominated. Gransand macropores provide minor contributions to sediment fabric. Porosities calculated for the matrix material correlate well whtotal sample porosities.

teetheihrn o

iacegis

(poened6 thn

rnthr

etiten;

theinsinig.e-hei-Co-lyn-

pe-13d

al.,y-–

260

di-X-arero-ym-reen-entde-s-

INTRODUCTION

Sites drilled during Ocean Drilling Program (ODP) Legs 160 and161 form an east–west, trans-Mediterranean transect focusedMediterranean paleoceanography as well as numerous regionaltonic objectives. The principal paleoceanographic objectives of L161 were to examine environmental conditions that prevailed in Western Mediterranean Sea during deposition of Pliocene–Pltocene sapropels in the eastern Mediterranean and to document tCenozoic history of hydrography and circulation within the westeMediterranean. Tectonic objectives of Leg 161 involved the originextensional basins in collisional settings and focused on the formtion of the Alboran Basin (Fig. 1).

Site 974 is located in a small basin within the central TyrrhenSea, a backarc basin that opened during late Miocene–Pleistotimes (Fig. 1). It is located ~300 m from Site 652, drilled during Le107 (Kastens, Mascle, Auroux, et al., 1987). The Pliocene–Pletocene sequence at Site 974 consists of hemipelagic sedimentsmarily nannofossil to nannofossil-rich clay and silty clay and nannfossil ooze), punctuated by organic-rich layers (lower organic contanalogs of the sapropels found throughout the eastern Mediterrane.g., Emeis, Robertson, Richter, et al., 1996) and volcaniclastic posits (McCoy and Cornell, 1990; Comas, Zahn, Klaus, et al., 199Site 975 is located on the Menorca Rise on the northern margin ofSouth Balearic Basin (Fig. 1). Nannofossil to calcareous clays asilty clays typify the Pliocene–Pleistocene hemipelagic sequencecovered at this site (Comas, Zahn, Klaus, et al., 1996). Documetion of organic-rich layers at Site 975 indicates paleoceanograpconditions associated with cyclic deposition of sapropel-like laye

1Zahn, R., Comas, M.C., and Klaus, A. (Eds.), 1999. Proc. ODP, Sci. Results, 161:College Station, TX (Ocean Drilling Program).

2 School of Ocean and Earth Science and Technology, University of Hawaii, Hono-lulu, Hawaii 96822, U.S.A. jtribble@soest.hawaii.edu

onc-

ges-e

fa-

nne

-ri-

-t

an;e-).ede-a-ics

extended farther to the west than previously thought (e.g., Emeisal., 1991). The Pliocene–Pleistocene hemipelagic sediments at S975 overlie micrites and evaporites of the latest Miocene (Messiniasee Marsaglia and Tribble, Chap. 1, this volume).

Four sites were drilled across the Alboran Sea to investigate subsidence histories and tectonic evolutions of the major sub-bas(the Western Alboran Basin [Site 976], the Eastern Alboran Bas[Sites 977 and 978], and the Southern Alboran Basin [Site 979]; F1). The Alboran Basin formed by extension of continental crust bhind arcuate thrust belts that remained active during evolution of tbasin; models explaining the origin of extension in an overall collsional setting are numerous and controversial (see discussion in mas, Zahn, Klaus, et al., 1996). The Alboran Basin has relativeshallow water depths (<2000 m) and contains up to 7 km of sedimetary fill (Comas et al., 1992; Jurado and Comas, 1992). The hemilagic sediments, sampled during Deep Sea Drilling Project Leg and ODP Leg 161, consist primarily of nannofossil-rich clays annannofossil to calcareous clays and silty clays (Ryan, Hsü, et 1973; Comas, Zahn, Klaus, et al., 1996). In addition to these cladominated lithologies, intervals of sand and or gravel of MiocenePliocene age were penetrated at Sites 976, 977, and 978. Nearlym of metamorphic basement was penetrated at Site 976.

This study addresses the long-term history of hemipelagic sementation in the western Mediterranean through quantitative bulk ray mineralogy analyses. In addition, the hemipelagic sediments characterized by backscattered electron microscope studies of micfabric. The microfabric of a sediment reflects its original mineralogand depositional environment, postdepositional diagenesis and copaction, and deformation. Changes in a sediment’s microfabric aaccompanied by changes in physical properties such as porosity, dsity, and acoustic velocity. In addition to providing a record of thcomposition and history of a sediment, microfabric is an importadetermining factor in the response of sediment to compaction and formation. Digital image analysis is used in this study to quantify a

99

J.S. TRIBBLE, R.H. WILKENS

5°W 0° 5° 10° 15° 20° 25°

35°

40°

45°

HellenicArc

Tyrrhenian

Ligurian

South Balearic

Pannonian

Site 975

Site 976

Alboran

Aegean

CalabrianArc

Site 974

Neogene oceanic crust

Neogene extended continental crust

Collision fronts

Alpine orogenic chains

Tell

Rif

Betic

Alps

Apennines

45°N

Site 979Site 977/8

Pyrénées

50°

GibraltarArc

Figure 1. Locations of Leg 161 sites in the Western Medi-terranean Sea.

g

i

mf

th

u

es,nateh in thanfactandardonasiteouldlay

ngesinlesidsx-le insam-e,ays then ac-heningentose

re-ric

eiss

nk-sery ise.lay-cesly-

pects of the sediment microfabric, and these data are related to sedi-ment composition and physical properties.

METHODS

Sample Selection

Samples were collected approximately once per section from onehole at each site drilled during Leg 161. During sample selection, em-phasis was placed on the dominant lithology of each core (generallyhemipelagite); only a few samples of minor lithologies were taken. Atotal of 857 samples were analyzed for major mineralogy. A subsetof 36 of these samples was also imaged with the scanning electronmicroscope (SEM). Hole 974B was chosen for detailed image analy-sis, whereas samples ~0.5, 2.0, and 5.0 Ma (nannozones 19F, 19A,and 12, respectively) were taken from each of the other sites so thatfabric of sediment deposited across the western Mediterranean dur-ing the same period of time could be compared. Almost all of thesamples were collected adjacent to a physical properties sample sothat imaging and mineralogy data could be compared to measure-ments of density, porosity, and velocity.

Mineralogy

X-ray diffraction (XRD) was used to determine bulk mineralogy.Samples from Holes 974B, 975B, 976B, 977A, and 978A were ana-lyzed on board ship according to methods described in ShipboardScientific Party (1996a). Samples from Hole 979A were analyzed atthe University of Hawaii (UH) using a Scintag PAD V powder dif-fractometer. The samples were continuously scanned from 2° to2θ, using a chopper increment of 0.03° and a scan speed of 1° 2θ/min.Unfiltered Cu Kα radiation was used with a tube voltage of 45 kand a tube current of 40 mA.

The relative abundances of total clay minerals, quartz, placlase, and calcite, normalized to 100%, were estimated for all bsamples using a routine developed at UH. The routine uses a staneous linear equation algorithm that accounts for peak ove(Karlak and Burnett, 1966) and was calibrated using a series of ticomponent mixtures as external standards. Details of the quantition method are given in Wilkens et al. (1992). Analysis of standmixtures of clay minerals, quartz, feldspar, and calcite using quantification routine indicate results are generally good to wit10% of the actual sample composition (Wilkens et al., 1992). Relution for these phases is on the order of 2% or better. To eval

100

70°

V

io-ulkmul-rlap

ul-ica-ardhisinso-ate

how well the quantification algorithm works on the Leg 161 samplXRD results for calcite and shipboard measurements of carbocontent (Comas, Zahn, Klaus, et al., 1996) are plotted vs. deptFigure 2. The geochemical analyses tend to be somewhat greaterthe XRD results; some of this difference may be a result of the that carbonate in the form of dolomite is present in the samples would have been included in the total carbonate of the shipboanalyses. Although dolomite was not included in the quantificatiroutine, the relative intensity of the main dolomite X-ray peak wdetermined. This relative intensity is simply a measure of dolompeak height relative to the most intense peak in the pattern and shnot be compared numerically with the weight percentages of cminerals, calcite, quartz, and feldspar.

Scanning Electron Microscopy

SEM samples were stored in sealed containers with damp spoto maintain fluid saturation during transport to the laboratory. Withseveral months of the cruise, 3- to 4-mm-thick slices of the sampchosen for SEM examination were treated to replace their pore fluwith epoxy. Replacement of the pore fluid involves two different echanges because the Spurr low-viscosity epoxy used is not solubwater. The exchange process was performed stepwise, soaking ples for 48 hr first in a fluid mixture of 80% water and 20% acetonand proceeding in 20% steps (60:40, 40:60, 20:80) every two duntil the sample was soaking in pure acetone. The process wasrepeated using an acetone-epoxy mixture, gradually replacing theetone in the sediment pores with epoxy. The samples were tcured, cut, polished using 0.25-µm grit, and mounted for study usa backscattered detector in the SEM. The efficacy of the replacemprocess was confirmed by the fact that internal voids such as thwithin foraminifer tests were epoxy filled. This process has beenpeatedly successful in preserving delicate sediment microfab(Tribble et al., 1993; Tribble and Wilkens, 1994).

Backscattered electron (BSE) images were collected on a ZDSM 962 SEM at five magnifications (100×, 200×, 500×, 1000×, and2000×). Digital files of these images are available in Appendix A oCD-ROM, this volume. In these digital images, the intensity of bacscattered electrons is directly proportional to density, so that denobjects appear brighter in the captured images, whereas epoxblack. Relatively large grains appear uniformly light gray to whitGrains are surrounded by matrix of a darker hue comprised of csized grains and microporosity, and relatively large pore spa(macropores), now filled with epoxy, are black. Digital image ana

SEDIMENT MINERALOGY AND MICROFABRIC

n theric Al- the Fig-at-

ar-wt%

ndheoles toTheentste inions00;er ofca-rm

er-site

D,ip-itesity-

sis was used to quantify the volumes occupied by these three fabricelements (grains, matrix, and macropores) using methods describedin Tribble and Wilkens (1994).

DESCRIPTION OF MINERALOGYAND MICROFABRIC

Hemipelagic Sediment

X-ray intensities of major phases are given in Appendix B (CD-ROM, this volume) for all samples analyzed, with the exception ofthe Miocene section from Site 975; those data are reported in Marsa-glia and Tribble (Chap. 1, this volume). X-ray mineralogy results aredisplayed as cumulative major mineral content vs. age up to 6 Ma inFigure 3. Because older sediments were recovered at Sites 976 and978, the entire age profiles for those sites are shown separately (Fig.4). Locations of SEM samples are indicated by circles on the figures.Ages were calculated by linear interpolation between biostratigraphicevent markers using data contained in individual site chapters (Co-mas, Zahn, Klaus, et al., 1996). In addition, average weight percent-ages of the four major phases, plus the average relative intensity ofthe dolomite peak, are given in Table 1 for each lithologic unit.

All of the mineralogy profiles are dominated by clay minerals andcalcite (Figs. 3, 4). Feldspar is present in only minor amounts in mostsamples. Quartz averages between 10% and 20% with a few excep-tions, most notably in samples near the Pliocene/Pleistocene bound-

200

150

100

50

0

De

pth

(m

bsf

)

0.80.60.40.20.0

Weight Fraction

Hole 974B Calcite by XRD CaCO3 by Coulometry

Figure 2. Depth profiles of carbonate content as measured by shipboard cou-lometric analysis and calcite content as determined by X-ray diffraction forsediment from Hole 974B. Overall agreement of the two data sets is quitegood. Offset to higher values of the shipboard carbonate analyses reflects inpart inclusion of noncalcite carbonates such as dolomite.

ary (2 Ma) at Sites 975 and 976, in the Pleistocene (0.4−0.7 Ma) atSite 974, and in sediments older than 6 Ma at Site 978. Figure 5 illus-trates a typical microfabric of the dominant hemipelagic sediments.Our backscattered electron micrographs were collected by imagingflat, polished surfaces; the images thus represent vertical cross sec-tions of sediment structure. The fabric of the hemipelagites is domi-nated by the medium gray matrix material (a mixture predominantlyof clay minerals and microporosity at a scale too fine for resolutionin this image; Fig. 5). Quantification of the fabric elements yieldedan average value of 91% matrix for this sample. Silt- to sand-sizedterrigenous grains (TG, primarily quartz with minor feldspar) andforaminifer fragments (F) appear as bright areas within the darkermatrix and total to an average of 9% grains. Most terrigenous grainsare <10 µm in diameter. No macropores are visible in this image sothe quantified macroporosity was negligible.

Calcite in the hemipelagic sediments is present primarily as bio-genic particles (primarily nannofossils and foraminifers). Abundantand well-preserved foraminifers are evident in images of many sam-ples, especially those from more eastern sites (e.g., Sample 161-974B-6H-7, 39 cm; Fig. 6). The foraminifer tests are generally notfilled with matrix material, a condition that tends to increase total po-rosity of the sediment. Interestingly, these tests often remain emptyof matrix material to depths of hundreds of meters and only begin tofill when recrystallization of calcite and biogenic silica begins (Wilk-ens et al., 1987). In Figure 6, some of the epoxy has been pluckedfrom inside the tests during polishing, but it appears that most testsare epoxy filled and did not contain matrix. This sample has a rela-tively high carbonate content (47% from XRD), as do many of thesamples below it, and is fairly typical of the hemipelagic sedimentsfrom Sites 974 and 975.

There is an east–west trend of decreasing calcite abundance ihemipelagic sediments (Fig. 3; Table 1), from ~40% in the Baleaand Tyrrhenian Seas, to ~20% calcite in the western Alboran Sea.though present, foraminifers are less abundant in sediments ofAlboran Sea, compared to the eastern sites (Sites 974 and 975).ure 7 illustrates the fabric of a typical western Alboran sample thcontains ~22 wt% calcite by XRD. Coccoliths, although not individually visible at the scale of this image, make up the bulk of the cbonate content. This sample also contains abundant quartz (16 by XRD).

Calcite contents of western Alboran sediments (Holes 976B a979A) do not change appreciably with age (Fig. 3; Table 1). In tEastern Alboran Sea (Holes 977A and 978A), Balearic Sea (H975B), and the Tyrrhenian Sea (Hole 974B), calcite content tenddecrease with decreasing age of the sediment (Fig. 3; Table 1). greatest concentrations of carbonate occur at Site 974 in sedimbetween 3 and 4.5 m.y. old. The greater abundance of carbonathe older sediments does not appear to reflect higher concentratof foraminifer tests. We analyzed all of the Hole 974B images (~3see Appendix A) and observed no increase in the average numbgrains per image in the older samples. However, higher magnifition revealed a large component of carbonate in the matrix in the foof coccolith platelets (Fig. 8).

To obtain a better understanding of spatial and temporal diffences in calcite content, calcite fluxes were calculated for each (where data were available) according to the following:

flux = wt. fract. × sed. rt. × BD × (1 – por),

where the flux is calculated in g CaCO3/cm2/m.y., wt. fract. is theweight fraction of calcite in the sediment as determined by XRsed. rt. is the sedimentation rate in cm/m.y. calculated from shboard determinations of biostratigraphic events (see individual schapters in Comas, Zahn, Klaus, et al., 1996), BD is the bulk denof each sample in g/cm3, and (1 – por) is the fraction of grains, cal

101

J.S. TRIBBLE, R.H. WILKENS

0.80.40.0

9 7 5 B

0.80.40.0

9 7 6 B

0.80.40.0

9 7 7 A

0.80.40.0

9 7 8 A

0.80.40.0

6

5

4

3

2

1

0A

ge

(M

a)

9 7 9 A

Quartz Calcite Clay Feldspar Images

0.80.40.0

9 7 4 B

Cumulative Total

Figure 3. Cumulative plots of bulk mineralogy vs. age up to 6 Ma for each of the Leg 161 sites. Location of microfabric samples are indicated by circles on theplots.

rlcf svlitorati

of iso-975re,raller

erro-

e-b-tesentte

itethepro-they- in-

culated as 1 minus the fractional porosity. Bulk density and porositydata are available for samples taken adjacent to nearly all of oursamples (Comas, Zahn, Klaus, et al., 1996). The resulting flux data(Fig. 9) represent the net burial of calcite and thereby reflect the bal-ance of delivery of calcite to the site and possible dissolution of cal-cite. Periods of high flux may represent times of relatively high cal-cite productivity in surface waters, times of general high sedimenta-tion rate during which detrital carbonate is delivered to the site,times of particularly low calcite dissolution, or any combinationthereof.

A generally west–east trend of decreasing calcite flux is appa(Fig. 9). Note that this trend is opposite to that defined by the cacontent data (Fig. 3; Table 1). Interestingly, the average calcite data for each site are correlated with current water depths of the(Table 2). This is true whether all flux data from each site are aaged, or whether a background flux average is determined by enating peaks from the calculation. Although Mediterranean botwaters are currently supersaturated with respect to calcite and anite (Millero et al., 1979; Takahashi et al., 1980), dissolution of ccite within the sediment column driven by organic matter degrada

102

entiteluxiteser-mi-m

ago-l-on

is possible (e.g., Archer et al., 1989). Background concentrationstotal organic carbon for the easternmost sites are low, despite thelated high values found in organic-rich layers (see Sites 974 and chapters in Comas, Zahn, Klaus, et al., 1996). It is likely, therefothat dissolution of carbonate plays at most a minor role. The ovewest–east trend of decreasing calcite flux more likely reflects highdelivery rates of calcite to the westernmost sites.

It is not possible from the XRD data alone to determine wheththe calcite flux values indicate a trend of decreasing carbonate pductivity from west to east, or whether the flux values primarily rflect detrital input rates. If the detrital sediment delivered to the Aloran Sea contains a significant fraction of carbonate, calcite flux rawould be high, although the percentage of calcite in the sedimcould still be relatively low because of dilution by the noncarbonafraction of the detrital sediment. The inverse correlation of calcflux with water depth across the six sites (Table 2) may reflect greater distance of deeper sites from shallow shelves that could vide a source of detrital carbonate. Our microfabric images of western Alboran sediments typically show abundant silt- to clasized grains, many of which appear from backscattered electron

SEDIMENT MINERALOGY AND MICROFABRIC

tese).z, toac-dy,or-

ngeanditeuc-

g.tyl.,ncetybal

n, itliv-lses

rmteslethepleancens.

tensities to be carbonate (e.g., Fig. 10). These grains are generallyrounded and show no obvious biogenic morphology or structure,which indicates a likely detrital origin.

The age profiles of calcite flux (Fig. 9) indicate no overall tempo-ral change in flux, except at Site 977 where sediments younger than1.5 Ma are characterized by somewhat higher calcite fluxes than theolder sediments. Each profile, however, is characterized by peaks of

1.00.80.60.40.2

Cumulative Fraction

1 0

8

6

4

2

0

Ag

e (

Ma

)

Hole 976B Quartz Calcite Clay Feldspar

0.80.40.0

6

5

4

3

2

Hole 978A Quartz Calcite Clay Feldspar

Figure 4. Cumulative plots of bulk mineralogy vs. age for Holes 976B and978A.

calcite flux above relatively uniform background flux levels for eachsite. Temporal correlations of some of these peaks from site to site(Fig. 9) indicate the regional nature of the causal events, whether theywere high productivity events or times of increased detrital flux ofcalcite. Comparison of BSE images of high and low calcite flux sam-ples shows no consistent difference in the ratio of biogenic to detritalcarbonate. Temporal correlations of peaks in carbonate content of thefine-grained fraction (<20 µm) of samples from the Alboran Sea sihave also been noted by Skilbeck and Tribble (Chap. 7, this volumThe lack of similar correlation in abundance of fine detrital quartand the frequency of the peaks (roughly every 0.5 Ma), lead thempostulate a climatic control on carbonate productivity as a causal ftor. The magnitudes of the calcite flux peaks calculated in this stuhowever, suggest that detrital input of carbonate to the sites is imptant. The background flux levels are comparable to the upper raof open ocean values of calcite mass accumulation rates (Farrell Prell, 1991). To accredit the order-of-magnitude increases in calcflux observed in the Alboran and Balearic Seas to increased prodtivity alone would be difficult.

The ~0.4- to 0.5-m.y. frequency of the peaks in calcite flux (Fi9) is similar to cyclicity observed in marine carbonates from a varieof settings, generally attributed to dissolution (e.g., Moore et a1982; Droxler et al., 1990; Bassinot et al., 1994). The correspondeof this frequency with the 0.413-m.y. period of Earth’s eccentricicycle has been noted, and a direct or indirect connection with gloclimate inferred, by these authors.

Although the role of dissolution in production of the westerMediterranean carbonate cycles cannot be totally discountedseems likely that the primary causative factors were increased deery of detrital carbonate to the basins, perhaps accompanied by puin productivity.

Quartz abundances of hemipelagic sediments are fairly unifoacross all six sites (Fig. 3; Table 1), although western Alboran siare perhaps a little richer in quartz. In the Tyrrhenian Basin (Ho974B), spikes of high quartz and feldspar content characterize Pleistocene sediments (Fig. 3). Microfabric examination of a samcorresponding to one of these peaks showed no increase in abundof large terrigenous grains in images collected at low magnificatioAt higher magnification, however, abundant grains under 10 µm areapparent (Fig. 11). It is possible that these grains may have beenwindborne and that the quartz and feldspar peaks may have paleocli-matologic significance. The three largest peaks of quartz abundanceall correlate with spikes in feldspar content, but not all samples with

Table 1. Leg 161 bulk mineralogy: average values for lithologic units.

Note: Sites are in order from west to east. — = no data.

HoleLithologic

unitDepth(mbsf) Sediment type

Clay minerals (wt%)

Quartz (wt%)

Feldspar (wt%)

Calcite(wt%)

Dolomite (relative intensity)

976B I 0-362.1 Hemipelagite 66 15 1 18 11II 362.1-518.3 Sand and clay 40 40 2 18 47III 518.3-660.2 Hemipelagite 64 14 1 21 9IV 660.2-669.73 Basal sands 32 25 31 12 39

979A I 0-580.9 Hemipelagite 57 17 1 26 5

977A IA 0-417.4 Hemipelagite 61 13 1 25 9IB 417.4-490.8 Hemipelagite 53 14 1 32 6IC 490.8-532.9 Hemipelagite 49 10 1 40 3

978A IA 213.0-342.2 Hemipelagite 60 12 1 27 10IB 342.2-409.3 Hemipelagite 57 12 1 30 9IC 409.3-620.9 Hemipelagite 53 11 1 35 4II 620.9-630.67 Pebbles — — — — —III 630.67-694.3 Sand/silt/clay 66 23 2 9 9

975B I 0-305.2 Hemipelagite 48 12 1 40 4II 305.2-307.0 Micrite and silty clay 41 13 1 45 13

974B I 0-88.9 Hemipelagite 60 12 3 25 6II 88.9-199.55 Hemipelagite 51 7 1 41 2III 199.55-203.86 Clay/silt/sand 55 29 7 9 8

103

J.S. TRIBBLE, R.H. WILKENS

161 - 974B - 1H - 3, 81cm

100 µm

TG

F

TG

TG

FFigure 5. BSE image of Sample 161-974B-1H-3, 81 cm (3.81 mbsf), collected at 200×. Bright (more dense) objects are either foraminifer fragments (F) or terrigenous grains (TG). The bulk of the sample is the matrix of clay minerals and microporosity, with admixed nannofossils.

161 - 974B - 6H - 7, 39cm

200 µm

Figure 6. BSE image of a foraminifer-rich sample from the Tyrrhenian Sea (Sample 161-974B-6H-7, 39 cm; 53.89 mbsf; collected at 100×). This carbonate-rich sam-ple contains abundant, well-preserved foraminifers typi-cal of most of the sediments from Holes 974B and 975B.

high percentages of feldspar are enriched in quartz (Fig. 3). Severalof the feldspar spikes correspond to samples taken from ash layers asidentified visually by shipboard scientists (Shipboard Scientific Par-ty, 1996b, table 4). The elevated quartz contents of some of these lay-ers may reflect entrainment of detrital material during emplacementof the ash layers by gravity flow (a common depositional mechanismfor ash layers in the Tyrrhenian Sea; McCoy and Cornell, 1990). Oth-er than the spikes of high feldspar content at Hole 974B, feldspar con-centrations of hemipelagic sediments are uniformly low.

Dolomite is present at all sites (Table 1). XRD identification ofthis phase is tentative; some dolomite is iron-rich and may actually

104

be ankerite. Background levels of dolomite in hemipelagic sedimentstend to decrease from west to east, possibly in conjunction with thedecrease in sedimentation rates. A detrital source of much of the do-lomite was suggested by shipboard scientists (Comas, Zahn, Klaus,et al., 1996).

SAND-RICH INTERVALS

Core recovery of sand-rich intervals was limited, but 27 samplesof sandy sediment from Sites 976 and 978 were analyzed for bulk

SEDIMENT MINERALOGY AND MICROFABRIC

161 - 976B - 60X - 3, 28cm

100 µm

Figure 7. BSE image of a typical hemipelagic sediment from the western Alboran Sea (Sample 161-976B-60X-3, 28 cm; 556.41 mbsf; collected at 200×). Lower cal-cite content at this site relative to more eastern sites is evident in the lower abundance of foraminifers and nannofossils.

CC

C

161 - 974B - 17H - 6, 57cm

20 µm

Figure 8. BSE image of a hemipelagic sediment from the Tyrrhenian Sea (Sample 161-974B-17H-6, 57 cm; 157.07 mbsf; collected at 1000×). Coccolith plates (C) dominate the matrix of this carbonate-rich sample.

mineralogy. Pliocene sand-rich lithologies encountered at Site 976(lithostratigraphic Unit II; see Fig. 4, Table 1) are enriched in dolo-mite and quartz. The lower half of this unit also contains feldsparconcentrations above background levels. Calcite abundances in thissandy unit are similar to those of adjacent hemipelagic sediments,which indicates a likely detrital source for some of the carbonate. OneSEM sample was taken from the top of the lithostratigraphic Unit IIsandy interval in Hole 976B (Fig. 12). Note the abundance of silt-sized grains and the breakage of the foraminifers. Higher sedimenta-

tion rates at Site 976 led to deeper burial and greater compaction thannoted in a stratigraphically equivalent sample from the Balearic Ba-sin (Hole 975B; Fig. 13). Basal sediments at Site 976 are enriched inquartz, feldspar, and dolomite relative to the shallower hemipelagicsediments (Fig. 4; Table 1). In contrast to the sand-rich intervals ofSite 976, the sand-rich unit (lithostratigraphic Unit III) of Mioceneage at the base of Hole 978A is marked by an increase in terrigenoussediment (quartz, clay minerals, and feldspar) and dolomite at the ex-pense of calcite (Table 1).

105

J.S. TRIBBLE, R.H. WILKENS

Figure 9. Calcite flux plotted vs. age for all six Leg 161 sites, arranged from west to east. Gaps in the data reflect incomplete core recovery, lack of physicalproperties measurements, or incomplete sedimentation rate data, usually because of hiatuses in the section.

106

Table 2. Average calcite flux and sedimentation rate data.

Note: Average Pleistocene and upper Pliocene sedimentation rates are from individual site chapters (Comas, Zahn, Klaus, et al., 1996).

SiteWater depth

(m)

Average calcite flux

(kg/cm2/m.y.)

Average background flux

(kg/cm2/m.y.)

Average Pleistocene

sedimentation rate (m/m.y.)

Average upper Pliocene

sedimentation rate (m/m.y.)

976 1108 4.1 3.0 208 340979 1062 5.0 3.7 200 187978 1929 4.6 3.1 127 111977 1984 3.9 2.8 154 96975 2417 2.5 1.5 68 48.9974 3459 1.1 0.8 55 27.2

nda-geofe thes-onionun-

ASH LAYERS

Although our emphasis in this study was on analysis of dominantlithologies, we selected two ash layers from the Tyrrhenian Sea (Site974) for microfabric analysis. A sample of vitric ash (Sample 161-974B-2H-3, 76 cm; Fig. 14) contains abundant volcanic glass. Mostof the glass shards are still intact, although some of the grains appearto be undergoing corrosion. Note the near absence of biogenic car-bonate (XRD calcite content is 11%), probably because the relativelyrapid deposition of the glass swamped any pelagic carbonate contri-bution to the sediment composition. Alignment of the glass shards isevident; unfortunately, although orientation of all BSE samples rela-tive to vertical was recorded, this sample broke into several piecesduring impregnation, and orientation of the pieces is indeterminable.An altered volcanic ash sample from greater depth in Hole 974B wasalso imaged (Sample 161-974B-13-6, 83 cm; Fig. 15). In this samplethe glass has dissolved away leaving voids that are now epoxy-filled.Compare the shapes of the empty casts with glass grains illustrated inFigure 14. This fabric is similar to that noted in previous studies of

ash-rich sediments from the Nankai Trough and Hawaiian Arch(Tribble et al., 1993; Tribble and Wilkens, 1994). Maintenance oflarge, open pores at this burial depth (118 m below seafloor [bsf]) re-quires that the clay matrix has developed some strength and rigidity(or, less likely in this case, that the pores are overpressured).

QUANTIFICATION OF MICROFABRIC

Quantification of the fabric elements—grains, macropores, amatrix—yielded no significant depth trends and no strong correltions with either mineralogy or physical properties. The percentaof grains, generally <10%, is weakly correlated with the sum quartz + feldspar. Although foraminifers are visually striking in somsamples, the lack of correlation between abundance of grains andpercentage of calcite reflects the dominant contribution of nannofosils to the calcite component. Macropores make a minor contributito total porosity (<5% in most samples) and show no depth variatin abundance. The few samples with >5% macropores contain ab

SEDIMENT MINERALOGY AND MICROFABRIC

161 - 976B - 20X - 3, 95cm

20µm

Figure 10. BSE image of a silt-rich sediment from the western Alboran Sea (Sample 161-976B-20X-3, 95 cm; 179.05 mbsf; collected at 1000×). This sample contains abundant silt-sized detrital grains, the brightest of which are thought to be primarily calcite.

161 - 974B - 4H - 3, 11cm

20 µm

Figure 11. BSE image of Sample 161-974B-4H-3, 11 cm (28.61 mbsf), collected at 1000×. This sample, from one of the quartz peaks in Figure 3, has abun-dant fine silt- and clay-sized grains of quartz.

cal-lb- cal-

dant foraminifers, which indicates the contribution of intratest poros-ity. The hemipelagic sediments imaged are matrix dominated. Matrixporosity can be calculated from our quantitative data as the differencebetween total sample porosity and macroporosity divided by the frac-tional area occupied by matrix. Matrix porosity for samples fromHole 974B approximates well the total porosity of the samples asmeasured on board ship (Fig. 16). The observed decrease in porositywith burial depth reflects loss of microporosity within the matrix viacompaction, with virtually no contribution of loss of macropores.

CONCLUSIONS

The mineralogy of hemipelagic sediments from all sites is domi-nated by clay minerals and calcite. Quartz and feldspar concentra-tions in the hemipelagic sediment are relatively low. Calcite contentincreases somewhat with depth at all sites except the westernmosttwo (Sites 976 and 979). There is a west–east trend of increasingcite content of hemipelagic sediments from ~20% in the western Aoran Sea to ~40% in the Balearic and Tyrrhenian Seas, although

107

J.S. TRIBBLE, R.H. WILKENS

161 - 976B - 39X - 3, 54cm

200 µm

Figure 12. BSE image of Sample 161-976B-39X-3, 54 cm (361.24 mbsf), collected at 100×. Foraminifers and silt-sized quartz and calcite grains are abundant in this ~2.0-Ma sample from the western Alboran Sea. Many of the foraminifer tests are broken because of compac-tion during burial, but it is only after the tests were sub-stantially crushed that they were filled in with matrix material.

161 - 975B - 15H - 3, 30cm

200 µm

Figure 13. BSE image of Sample 161-975B-15H-3, 30 cm (131.94 mbsf), collected at 100×. Abundant, well-preserved foraminifers are present in this ~2.0-Ma sample from the Balearic Sea. Most foraminifer cham-bers are empty or contain minor amounts of matrix material. Compare to Figure 12.

culations of calcite flux indicate an opposite trend of decreasing cal-cite accumulation rates from west to east.

Although our sampling frequency is rather coarse (generally 3samples per core for bulk mineralogy) compared to most studies ofpaleoceanography, pulses of calcite flux (mass accumulation rate)have been identified at all western Mediterranean sites. The frequen-cy of these flux peaks (~0.4−0.5 m.y.) is similar to that of carbonate

108

dissolution cycles noted globally (e.g., Moore et al. 1982; Droxler etal., 1990; Bassinot et al., 1994), but the magnitude and character ofthe Mediterranean flux peaks differ from the fluctuations attributedto dissolution. The calcite flux peaks at the western Mediterraneansites are as much as an order of magnitude greater than backgroundflux levels. In addition, the calcite flux variations of this study appearas peaks above relatively constant background flux levels rather than

SEDIMENT MINERALOGY AND MICROFABRIC

GG

161 - 974B - 2H - 3, 76cm

100 µm

Figure 14. BSE image of a vitric ash sample from the Tyrrhenian Basin (Sample 161-974B-2H-3, 76 cm; 10.26 mbsf; collected at 200×). Grains of volcanic glass (G) show a net alignment.

161 - 974B - 13H - 6, 83cm

100 µm

Figure 15. BSE image of an altered ash sample from the Tyrrhenian Basin (Sample 161-974B-13H-6, 83 cm; 119.33 mbsf; collected at 200×). The glass grains of this altered ash have dissolved away, leaving shard-shaped pores. Some of these pores are partially filled by crystal growth, probably of zeolites.

ao

s

ir.

%

for-akestlled

inrtieset al.,la-

true “cycles.” Although a complete explanation of the patterns of ccite accumulation in the western Mediterranean is beyond the scof this study, it appears likely that influx of detrital carbonate playsignificant role. Potential variations in productivity may also be important. Climate change, with associated changes in runoff, nutrdelivery to the basins, and atmospheric dust fluxes, is the likely ding force behind both the detrital input and productivity variations

Microfabric of the hemipelagic sediments is dominated by a clamineral-rich matrix. Quantifiable grains generally constitute <10

l-pe

a-

entiv-

y-

of the area examined. Quartz and, to a lesser extent, feldspar andaminifers are the major types of grains identified. Macropores ma minor contribution to total sample porosity (generally <5%). Momacropores are associated with foraminifer tests that were not fiin by matrix material.

Although microfabric quantification has proven a useful tool understanding the ties between compositional and physical propechanges in sedimentary sequences in other studies (e.g., Ehrlich 1984; Ross and Ehrlich, 1991; Tribble and Wilkens, 1994), the re

109

J.S. TRIBBLE, R.H. WILKENS

.ndds

tro-

.,per

ic

ray

e-

in

ries: a

rs,

nts

est

heder-

a

o

t

tive homogeneity of the western Mediterranean hemipelagic se-quences yielded a notable lack of trends in the microfabric data. Thecorrelation between calculated matrix porosities and total sample po-rosities as measured on shipboard was the single exception and re-flects the dominance of the clay matrix.

ACKNOWLEDGMENTS

The authors would like to thank Marianne Cusimano for tirelesswork in the XRD laboratory of the JOIDES Resolution, Jack Kronenfor assistance in image collection, and Yonat Swimmer for dataanalysis. This work was supported by grants to both authors from theUnited States Science Advisory Committee.

REFERENCES

Archer, D., Emerson, S., and Reimers, C., 1989. Dissolution of calcite indeep-sea sediments: pH and O2 microelectrode results. Geochim. Cosmo-chim. Acta, 53:2831–2845.

Bassinot, F.C., Beaufort, L., Vincent, E., Labeyrie, L.D., Rosteck, F., MüllP.J., Quidelleur, X., and Lancelot, Y., 1994. Coarse fraction fluctuatioin pelagic carbonate sediments from the tropical Indian Ocean: a 1,kyr record of carbonate dissolution. Paleoceanography, 9:579–600.

Comas, M.C., García-Dueñas, V., and Jurado, M.J., 1992. Neogene tecevolution of the Alboran Basin from MCS data. Geo-Mar. Lett., 12:157–164.

200

150

100

50

De

pth

(m

bsf

)

0.80.70.60.50.4

Fractional Porosity

974B Measured Porosity Image Porosity

Figure 16. Depth trends of shipboard total porosity measurements andcalculated values for matrix porosity derived from quantitative image ansis. Shipboard porosity was measured on samples immediately adjacethe microfabric samples. The close correspondence of the matrix pordata and the total porosity measurements reflects the minor contributionlarge pore space (macropores) makes to total porosity in these sedimen

110

er,ns500

tonic

Comas, M.C., Zahn, R., Klaus, A., et al., 1996. Proc. ODP, Init. Repts., 161:College Station, TX (Ocean Drilling Program).

Droxler, A.W., Haddad, G.A., Mucciarone, D.A., and Cullen, J.L., 1990Pliocene–Pleistocene aragonite cyclic variations in Holes 714A a716B (The Maldives) compared with Hole 633A (The Bahamas): recorof climate-induced CaCO3 preservation at intermediate water depths. InDuncan, R.A., Backman, J., Peterson, L.C., et al., Proc. ODP, Sci.Results, 115: College Station, TX (Ocean Drilling Program), 539–577.

Ehrlich, R., Crabtree, S.J., Kennedy, S.K., and Cannon, R.L., 1984. Pegraphic image analysis 1, analysis of reservoir complexes. J. Sediment.Petrol., 54:1365–1376.

Emeis, K.-C., Camerlenghi, A., McKenzie, J.A., Rio, D., and Sprovieri, R1991. The occurrence and significance of Pleistocene and UpPliocene sapropels in the Tyrrhenian Sea. Mar. Geol., 100:155–182.

Emeis, K.-C., Robertson, A.H.F., Richter, C., et al., 1996. Proc. ODP, Init.Repts., 160: College Station, TX (Ocean Drilling Program).

Farrell, J.W., and Prell, W.L., 1991. Pacific CaCO3 preservation and δ18Osince 4 Ma: paleoceanic and paleoclimatic implications. Paleoceanogra-phy, 6:485–498.

Jurado, M.J., and Comas, M.C., 1992. Well log interpretation and seismcharacter of the Cenozoic sequence in the Northern Alboran Sea. Geo-Mar. Lett., 12:129–136.

Karlak, R.F., and Burnett, D.S., 1966. Quantitative phase analysis by X-diffraction. Anal. Chem., 38:1741–1745.

Kastens, K.A., Mascle, J., Auroux, C., et al., 1987. Proc. ODP, Init. Repts.,107: College Station, TX (Ocean Drilling Program).

McCoy, F.W., and Cornell, W., 1990. Volcaniclastic sediments in the Tyrrhnian Basin. In Kastens, K.A., Mascle, J., et al., Proc. ODP, Sci. Results,107: College Station, TX (Ocean Drilling Program), 291–305.

Millero, F.J., Morse, J.W., and Chen, C.-T., 1979. The carbonate systemthe western Mediterranean Sea. Deep-Sea Res., 26A:1395–1404.

Moore, T.C., Jr., Pisias, N.G., and Dunn, D.A., 1982. Carbonate time seof the Quaternary and late Miocene sediments in the Pacific Oceanspectral comparison. Mar. Geol., 46:217–233.

Ross, C.M., and Ehrlich, R., 1991. Objective Measurement and Classifica-tion of Microfabrics and their Relationship to Physical Properties: NewYork (Springer-Verlag).

Ryan, W.B.F., Hsü, K.J., et al., 1973. Init. Repts. DSDP, 13 (Pts. 1 and 2):Washington (U.S. Govt. Printing Office).

Shipboard Scientific Party, 1996a. Explanatory notes. In Comas, M.C., Zahn,R., Klaus, A., et al., Proc. ODP, Init. Repts., 161: College Station, TX(Ocean Drilling Program), 21–49.

Shipboard Scientific Party, 1996b. Site 974. In Comas, M.C., Zahn, R.,Klaus, A., et al., Proc. ODP, Init. Repts., 161: College Station, TX(Ocean Drilling Program), 55–111.

Takahashi, T., Broecker, W.S., Bainbridge, A.E., and Weiss, R.F., 1980. Car-bonate Chemistry of the Atlantic, Pacific, and Indian Oceans: TheResults of the GEOSEC Expeditions, 1972–1978. Lamont-Doherty Geol.Observ. Tech. Rep. 1, CV-1–80.

Tribble, J.S., and Wilkens, R.H., 1994. Microfabric of altered ash layeODP Leg 131, Nankai Trough. Clays Clay Miner., 42:428–436.

Tribble, J.S., Wilkens, R., Arvidson, R.S., and Busing, C.J., 1993. Sedimeof the Hawaiian Arch: X-ray mineralogy and microfabric. In Wilkens,R.H., Firth, J., Bender, J., et al., Proc. ODP, Sci. Results, 136: CollegeStation, TX (Ocean Drilling Program), 65–76.

Wilkens, R.H., De Carlo, E.H., and Tribble, J.S., 1992. Data report: X-raybulk mineralogy of Exmouth and Wombat Plateau sediments, northwAustralian margin. In von Rad, U., Haq, B.U., et al., Proc. ODP, Sci.Results, 122: College Station, TX (Ocean Drilling Program), 887–896.

Wilkens, R.H., Schreiber, B.C., Caruso, L., and Simmons, G., 1987. Teffects of diagenesis on the microstructure of Eocene sediments boring the Baltimore Canyon Trough. In Poag, C.W., Watts, A.B., et al., Init.Repts. DSDP, 95: Washington (U.S. Govt. Printing Office), 527–547.

Date of initial receipt: 7 May 1997Date of acceptance: 26 January 1998Ms 161SR-252

thely-nt tosity

thats.